Add option for using new default kwargs

Author: jsignell

xarray/backends/api.py

@@ -1402,14 +1402,14 @@ def open_mfdataset(
         | Sequence[Index]
         | None
     ) = None,
-    compat: CompatOptions = "no_conflicts",
+    compat: CompatOptions | None = None,
     preprocess: Callable[[Dataset], Dataset] | None = None,
     engine: T_Engine | None = None,
-    data_vars: Literal["all", "minimal", "different"] | list[str] = "all",
-    coords="different",
+    data_vars: Literal["all", "minimal", "different"] | list[str] | None = None,
+    coords=None,
     combine: Literal["by_coords", "nested"] = "by_coords",
     parallel: bool = False,
-    join: JoinOptions = "outer",
+    join: JoinOptions | None = None,
     attrs_file: str | os.PathLike | None = None,
     combine_attrs: CombineAttrsOptions = "override",
     **kwargs,
@@ -1596,9 +1596,6 @@ def open_mfdataset(
 
     paths1d: list[str | ReadBuffer]
     if combine == "nested":
-        if isinstance(concat_dim, str | DataArray) or concat_dim is None:
-            concat_dim = [concat_dim]  # type: ignore[assignment]
-
         # This creates a flat list which is easier to iterate over, whilst
         # encoding the originally-supplied structure as "ids".
         # The "ids" are not used at all if combine='by_coords`.
@@ -1647,7 +1644,7 @@ def open_mfdataset(
             # along each dimension, using structure given by "ids"
             combined = _nested_combine(
                 datasets,
-                concat_dims=concat_dim,
+                concat_dim=concat_dim,
                 compat=compat,
                 data_vars=data_vars,
                 coords=coords,

xarray/core/alignment.py

@@ -19,6 +19,7 @@
     indexes_all_equal,
     safe_cast_to_index,
 )
+from xarray.core.options import _new_default_combine_kwargs_warning
 from xarray.core.types import T_Alignable
 from xarray.core.utils import is_dict_like, is_full_slice
 from xarray.core.variable import Variable, as_compatible_data, calculate_dimensions
@@ -113,6 +114,7 @@ class Aligner(Generic[T_Alignable]):
     results: tuple[T_Alignable, ...]
     objects_matching_indexes: tuple[dict[MatchingIndexKey, Index], ...]
     join: str
+    join_is_default: bool = False
     exclude_dims: frozenset[Hashable]
     exclude_vars: frozenset[Hashable]
     copy: bool
@@ -418,12 +420,31 @@ def align_indexes(self) -> None:
                 else:
                     need_reindex = False
                 if need_reindex:
+                    if self.join_is_default and self.join != "exact":
+                        _new_default_combine_kwargs_warning(
+                            "join",
+                            self.join,
+                            "This change will result in the following ValueError:"
+                            "cannot be aligned with join='exact' because "
+                            "index/labels/sizes are not equal along "
+                            "these coordinates (dimensions): "
+                            + ", ".join(f"{name!r} {dims!r}" for name, dims in key[0]),
+                            recommend_set_options=False,
+                        )
                     if self.join == "exact":
+                        new_default_warning = (
+                            "Failure might be related to new default (join='exact'). "
+                            "Previously the default was join='outer'. "
+                            "The recommendation is to set join explicitly for this case."
+                        )
                         raise ValueError(
                             "cannot align objects with join='exact' where "
                             "index/labels/sizes are not equal along "
                             "these coordinates (dimensions): "
                             + ", ".join(f"{name!r} {dims!r}" for name, dims in key[0])
+                            + new_default_warning
+                            if self.join_is_default
+                            else ""
                         )
                     joiner = self._get_index_joiner(index_cls)
                     joined_index = joiner(matching_indexes)
@@ -892,6 +913,7 @@ def deep_align(
     exclude: str | Iterable[Hashable] = frozenset(),
     raise_on_invalid: bool = True,
     fill_value=dtypes.NA,
+    join_is_default: bool = False,
 ) -> list[Any]:
     """Align objects for merging, recursing into dictionary values.
 
@@ -944,14 +966,17 @@ def is_alignable(obj):
         else:
             out.append(variables)
 
-    aligned = align(
-        *targets,
+    aligner = Aligner(
+        targets,
         join=join,
         copy=copy,
         indexes=indexes,
-        exclude=exclude,
+        exclude_dims=exclude,
         fill_value=fill_value,
     )
+    aligner.join_is_default = join_is_default
+    aligner.align()
+    aligned = list(aligner.results)
 
     for position, key, aligned_obj in zip(positions, keys, aligned, strict=True):
         if key is no_key:

xarray/core/combine.py

@@ -202,9 +202,9 @@ def _combine_nd(
     concat_dims,
     data_vars,
     coords,
-    compat: CompatOptions,
+    compat: CompatOptions | None,
     fill_value,
-    join: JoinOptions,
+    join: JoinOptions | None,
     combine_attrs: CombineAttrsOptions,
 ):
     """
@@ -314,6 +314,7 @@ def _combine_1d(
                 fill_value=fill_value,
                 join=join,
                 combine_attrs=combine_attrs,
+                _from_concat=False,
             )
         except ValueError as err:
             if "encountered unexpected variable" in str(err):
@@ -345,18 +346,21 @@ def _new_tile_id(single_id_ds_pair):
 
 def _nested_combine(
     datasets,
-    concat_dims,
+    concat_dim,
     compat,
     data_vars,
     coords,
     ids,
     fill_value,
-    join: JoinOptions,
+    join,
     combine_attrs: CombineAttrsOptions,
 ):
     if len(datasets) == 0:
         return Dataset()
 
+    if isinstance(concat_dim, str | DataArray) or concat_dim is None:
+        concat_dim = [concat_dim]  # type: ignore[assignment]
+
     # Arrange datasets for concatenation
     # Use information from the shape of the user input
     if not ids:
@@ -373,10 +377,10 @@ def _nested_combine(
     # Apply series of concatenate or merge operations along each dimension
     combined = _combine_nd(
         combined_ids,
-        concat_dims,
-        compat=compat,
+        concat_dims=concat_dim,
         data_vars=data_vars,
         coords=coords,
+        compat=compat,
         fill_value=fill_value,
         join=join,
         combine_attrs=combine_attrs,
@@ -391,11 +395,11 @@ def _nested_combine(
 def combine_nested(
     datasets: DATASET_HYPERCUBE,
     concat_dim: str | DataArray | None | Sequence[str | DataArray | pd.Index | None],
-    compat: str = "no_conflicts",
-    data_vars: str = "all",
-    coords: str = "different",
+    compat: str | None = None,
+    data_vars: str | None = None,
+    coords: str | None = None,
     fill_value: object = dtypes.NA,
-    join: JoinOptions = "outer",
+    join: JoinOptions | None = None,
     combine_attrs: CombineAttrsOptions = "drop",
 ) -> Dataset:
     """
@@ -588,13 +592,10 @@ def combine_nested(
     if mixed_datasets_and_arrays:
         raise ValueError("Can't combine datasets with unnamed arrays.")
 
-    if isinstance(concat_dim, str | DataArray) or concat_dim is None:
-        concat_dim = [concat_dim]
-
     # The IDs argument tells _nested_combine that datasets aren't yet sorted
     return _nested_combine(
         datasets,
-        concat_dims=concat_dim,
+        concat_dim=concat_dim,
         compat=compat,
         data_vars=data_vars,
         coords=coords,
@@ -629,8 +630,8 @@ def _combine_single_variable_hypercube(
     fill_value,
     data_vars,
     coords,
-    compat: CompatOptions,
-    join: JoinOptions,
+    compat: CompatOptions | None,
+    join: JoinOptions | None,
     combine_attrs: CombineAttrsOptions,
 ):
     """
@@ -685,11 +686,11 @@ def _combine_single_variable_hypercube(
 
 def combine_by_coords(
     data_objects: Iterable[Dataset | DataArray] = [],
-    compat: CompatOptions = "no_conflicts",
-    data_vars: Literal["all", "minimal", "different"] | list[str] = "all",
-    coords: str = "different",
+    compat: CompatOptions | None = None,
+    data_vars: Literal["all", "minimal", "different"] | list[str] | None = None,
+    coords: str | None = None,
     fill_value: object = dtypes.NA,
-    join: JoinOptions = "outer",
+    join: JoinOptions | None = None,
     combine_attrs: CombineAttrsOptions = "no_conflicts",
 ) -> Dataset | DataArray:
     """